Primary biliary cholangitis (PBC) is a chronic cholestatic disorder characterized by progressive, non-suppurative inflammation and destruction of intrahepatic small bile ducts. The presence of antimithocondrial antibody (AMA) in the sera is a very important finding for the diagnosis of PBC. The sensitivity and specificity of AMA for PBC are both >90%–95%; therefore, AMA is the major hallmark of PBC . However, it has been reported that 5%–10% of patients with clinical, biochemical and histological features compatible with PBC do not have detectable AMA. In these patients, AMA-negative results obtained by indirect immunofluorescence (IIF) were confirmed by immunoblotting or ELISA .
These AMA-negative PBC patients seem to be different from positive ones with respect to production of other antibodies . The identification of reliable diagnostic biomarkers for AMA-negative PBC is necessary and some potential biomarkers are under consideration.
Several types of antinuclear antibodies (ANA), detected by IIF on HEp-2 cells, have been associated with PBC .
Particularly, within the spectrum of ANA staining patterns, the “multiple nuclear dots” (MND) is specific for PBC. The MND pattern is the result of antibodies directed against structural components of premyelocytic leukemia protein (PML) and Sp100-containing nuclear bodies (NBs). Another ANA pattern of particular interest among PBC patients is the rim-like/membranous pattern in which antibodies are directed against several proteins (about 80–100 proteins), specifically Gp210 antigen.
The detection of Sp100 and Gp210 antibodies by commercially available methods, such as enzyme-linked immunosorbent assays (ELISA) or immunoblot, could represent a further approach to complete ANA evaluation .
Finally, two novel PBC autoantigens, kelch-like 12 (KLHL12) and hexokinase 1 (HK1), identified by proteomic strategy, have been discovered .
The aim of this review was to analyze studies that evaluated the performance of serum markers in the diagnosis of AMA-negative PBC.
Materials and methods
Data sources and study selection
A systematic literature review was conducted and studies published only in English were identified searching in PubMed, Embase and Scopus and using the following keywords: “primary biliary cholangitis”, “primary biliary cirrhosis”, “antimitochondrial antibodies-negative”, “anti-mitochondrial antibodies-negative”, “anti-mitochondrial antibody-negative”, “antimitochondrial antibody-negative”, “ama-negative”. All abstracts were scanned, and studies evaluating the utility of assaying different serum markers for diagnosis of PBC and providing data on sensitivity and specificity were included.
Reviews, case reports, letters, conference abstracts and editorials were not included.
Data extraction and study quality assessment
The name of the first author, year of publication, country of origin, methods of antibody testing, sample size, and serum markers were extracted.
Where data were available, 2×2 tables were constructed to derive sensitivity, specificity, predictive value (PPV and NPV), likelihood ratio (LR+ and LR−) and odds ratio (OR).
Data analysis was performed using Microsoft Office Excel. Sensitivity, specificity, PPV, NPV, LR+ and LR− were obtained using http://www.sipmel.it/it/risorse/strumenti/calcolatore.
OR were calculated using true-positive (TP), false-positive (FP), false-negative (FN) and true-negative (TN) data.
Characteristics of included studies
The search strategy retrieved 370 potentially relevant studies. After reading titles and abstracts, 362 citations were excluded according to the selection criteria. According to the inclusion criteria, eight full-text studies were included in this meta-analysis.
The baseline characteristics of all included studies are summarized in Table 1. In particular, this table reports the following for each study: the first author of the study, the year of publication, where the study was conducted, the study design (retrospective or prospective), the serum biomarkers investigated, the assay methods used to measure them, the number of patients (with PBC and negative for AMA) and controls, which could be healthy subjects or with other diseases.
Table 2 summarized the number of true positive, false positive, false negative and true negative according to the result obtained for each serum biomarker measured with each assay method of each study. In particular, AMA-negative PBC patients were considered as true positive or false negative when the serum biomarker resulted positive or negative, respectively. By contrast, control patients were classified as false positive or true negative when the serum biomarker resulted positive or negative, respectively. Also, AMA-positive PBC patients were classified as controls.
The positive and negative LRs mean values were 29.15 (95% IC 10.41–93.72) and 0.76 (95% IC 0.55–1.08), respectively. The pooled sensitivity and specificity mean values were 37% (min–max, 8%–71%) and 85% (min–max, 47%–100%), respectively. The overall PPV and NPV mean values were 45% (min–max, 12%–100%) and 83% (min–max, 50%–99%). Finally, pooled OR was 59.33 (0.26–848.6).
Five studies had used the IIF method, five had used immunoblot technique and four had used ELISA method. The sensitivities obtained with IIF ranged from 17% to 71%, whereas specificities ranged from 69% to 100%. The sensitivities obtained by ELISA methods ranged from 15% to 55% and specificities ranged from 47% to 100%. Sensitivities and specificities by immunoblot ranged from 8% to 65% and 57% to 100%, respectively.
ORs obtained with the three methods were 160.10 (IIF), 19.08 (ELISA) and 23.88 (immunoblot).
The aim of this study was to evaluate the diagnostic value of several antibodies, detected with different methods, in AMA-negative PBC. All PBC patients had a diagnosis of PBC according to internationally accepted criteria (AASLD) and almost all had a liver biopsy-proven PBC .
Generally, PBC is characterized by the presence of anti-mitochondrial antibodies (AMA), which is the diagnostic hallmark of the disease, occurring in about 90%–95% of patients .
The immunological profile of PBC includes ANA detection (by IIF on HEp-2 cells; patterns PBC-related: MND and rim-like) in up to 50%–70% of cases . Moreover, antibodies to Sp100 and PML (related to MND pattern) and Gp210 (related to rim-like pattern), detected by ELISA or immunoblot, have potential diagnostic role in AMA-negative PBC.
Eight/eight selected studies reported poor diagnostic sensitivity (37%) but elevated specificity (85%) for all serum biomarkers. In particular, specificity was higher in AMA-negative PBC patients compared with non-PBC controls; when the comparison was with AMA-positive PBC patients, the value of specificity was sometimes more modest (<60%). Furthermore, the simultaneous detection of more biomarkers occurred only in cases with PBC; in particular, ANA reactivities (MND and rim-like patterns) were more frequently detected in AMA-negative than AMA-positive PBC.
Considering each method, the best diagnostic accuracy was obtained with IIF method, which had values of sensitivity higher (mean value 46%) compared to other methods.
Granito et al.  evaluated several antibodies in PBC patients with and without AMA antibodies, such as anti-Sp140, anti-Sp100 and anti-PML; although the diagnostic sensitivity of each antibody was not suitable as a good diagnostic marker, AMA-negative PBC patients had significantly more anti-Sp140/Sp100/PML (p=0.04) or anti-Sp140/anti-Sp100 (p=0.0003) and at least one of the three anti-NB reactivities (p=0.0003) than their AMA-positive counterpart. Their findings strengthened a potential diagnostic role of anti-NBs reactivities, especially when AMA are absent.
As reported by Muratori et al. , the overall prevalence of ANA was significantly higher in AMA-negative than in AMA-positive PBC patients; particularly, the MND and the rim-like pattern were more often observed in AMA-negative PBC patients (p<0.05 and p<0.01, respectively). A similar result was observed also for the positivity of anti-Sp100 more frequently recorded in AMA-negative patients.
Some studies , ,  demonstrated a more frequent detection of the aforementioned markers, alone or in combination (MND and rim-like pattern or anti-Sp100, anti-Sp140 or anti-Gp210) in AMA-negative PBC patients compared to those positive for AMA antibodies, acquiring a potential significance as positive serologic markers for these patients.
Granito et al.  reported a similar prevalence of the two ANA pattern (MND and rim-like) to that of the two corresponding ELISA reactivities anti-Sp100 and anti-Gp210 in AMA-negative PBC. These observation was often shared by other studies such as that of Liu et al.  and Muratori et al. .
Besides, the simultaneous positivity for both these antibodies (anti-Sp100 and anti-Gp210), like that for both the ANA patterns (MND and rim-like), was observed only in PBC cases (100% PPV), irrespective the AMA status. On the contrary, positivity for MND or rim-like patterns, not confirmed by ELISA method (anti-Sp100 and anti-Gp210), were found also in non-PBC patients (mainly in SLE). Granito et al.  concluded that the positivity for only MND or rim-like pattern, without AMA and anti-Sp100 or anti-Gp210 (by ELISA), could allow to rule out PBC diagnosis with more confidence.
Invernizzi et al.  reported a significant difference in prevalence of ANA positivity in AMA-negative PBC patients compared to AMA-positive patients (p=0.0002), but they did not specify the ANA pattern.
In fact, this is the only study in which sensitivity (71%) was higher than specificity (69%); probably, these results would have been different if they had considered the patterns (MND and rim-like) specific for PBC.
Regarding antibodies to CA II, sensitivity was always very low and this result ruled out the possible usefulness of these antibodies as diagnostic marker for AMA-negative PBC. Both studies reported these antibodies in patients with autoimmune liver disease, but there is no specific association with AMA-negative PBC. Therefore, anti-CA II could be considered a nonspecific marker of autoimmunity rather than a distinctive feature of AMA-negative PBC , .
More recently, new non-invasive PBC biomarkers, antibodies to KLHL12 and hexokinase-1 (HK1), have been identified by proteome microarrays . Both these autoantibodies had high specificity for PBC (>95%) compared with all non-PBC disease controls. Authors suggested the addition of these new analytes (by immunoblot/ELISA) to serological assessment, along with AMA and ANA assays (by IIF), to improve the overall sensitivity. Notably, these autoantibodies could be more useful in AMA-negative PBC, in combination with ANA, anti-GP210 and anti-Sp100. Even if there are no other studies that evaluate the diagnostic value of these biomarkers, we decided to cite this article, considering the importance of an accurate diagnosis of PBC at early stages, in particular for AMA-negative PBC patients, in order to start an early treatment, delaying liver failure and improving the survival rate of PBC.
All studies reported no significant differences between AMA-positive and AMA-negative PBC patients in terms of clinical and biochemical features (i.e. alkaline phosphatase, aminotransferase, albumin, bilirubin, etc.), the only differences were confined to the serological expression of non-organ specific autoantibodies, suggesting that the presence of AMA has diagnostic significance but does not influence the clinical course of patients with PBC .
OR confirms the “goodness” of these markers which are very specific for the PBC but the identification of effective diagnostic biomarkers for AMA-negative PBC patients is still needed, because antibodies considered in all studied are not as sensitive as AMA.
Granito A, Yang WH, Muratori L, Lim MJ, Nakajima A, Ferri S, et al. PML nuclear body component sp140 is a novel autoantigen in primary biliary cirrhosis. Am J Gastroenterol 2010;105:125–31. CrossrefWeb of SciencePubMedGoogle Scholar
Liu B, Shi XH, Zhang FC, Zhang W, Gao LX. Antimitochondrial antibody-negative primary biliary cirrhosis: a subset of primary biliary cirrhosis. Liver Int 2008;28:233–9. PubMedWeb of ScienceCrossrefGoogle Scholar
Granito A, Muratori P, Muratori L, Pappas G, Cassani F, Worthington J, et al. Antinuclear antibodies giving the ‘multiple nuclear dots’ or the ‘rim-like/membranous’ patterns: diagnostic accuracy for primary biliary cirrhosis. Aliment Pharmacol Ther 2006;24:1575–83. CrossrefPubMedGoogle Scholar
Ueno Y, Ishii M, Takahashi S, Mano Y, Yahagi K, Kisara N, et al. Primary biliary cirrhosis with antibody against carbonic anhydrase II associates with distinct immunological backgrounds. Hepatol Res 2001;20:18–27. CrossrefPubMedGoogle Scholar
Gordon SC, Quattrociocchi-Longe TM, Khan BA, Kodali VP, Chen J, Silverman AL, et al. Antibodies to carbonic anhydrase in patients with immune cholangiopathies. Gastroenterology 1995;108:1802–9. CrossrefPubMedGoogle Scholar
Norman GL, Yang CY, Ostendorff HP, Shums Z, Lim MJ, Wang J, et al. Anti-kelch-like 12 and anti-hexokinase 1: novel autoantibodies in primary biliary cirrhosis. Liver Int 2015;35:642–51. PubMedWeb of ScienceCrossrefGoogle Scholar
Moher D, Liberati A, Tetzlaff J, Altman DG. The PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med 2009;6:e1000097. CrossrefPubMedGoogle Scholar
Heathcote EJ. Management of primary biliary cirrhosis. The American Association for the study of liver disease practice guidelines. Hepatology 2000;31:1005–13. Google Scholar
Rigopoulou EI, Davies ET, Pares A, Zachou K, Liaskos C, Bogdanos DP, et al. Prevalence and clinical significance of isotype specific antinuclear antibodies in primary biliary cirrhosis. Gut 2005;54:528–32. PubMedCrossrefGoogle Scholar
Muratori P, Muratori L, Ferrari R, Cassani F, Bianchi G, Lenzi M, et al. Characterization and clinical impact of antinuclear antibodies in primary biliary cirrhosis. Am J Gastroenterol 2003;98:431–7. PubMedCrossrefGoogle Scholar
Invernizzi P, Crosignani A, Battezzati PM, Covini G, De Valle G, Larghi A, et al. Comparison of the clinical features and clinical course of antimitochondrial antibody-positive and -negative primary biliary cirrhosis. Hepatology 1997;25:1090–5. CrossrefPubMedGoogle Scholar
Comay D, Cauch-Dudek K, Hemphill D, Diamandis E, Wanless I, Heathcote EJ. Are antibodies to carbonic anhydrase II specific for anti-mitochondrial antibody-negative primary biliary cirrhosis? Dig Dis Sci 2000;45:2018–21. PubMedCrossrefGoogle Scholar
Invernizzi P, Battezzati PM, Crosignani A, Zermiani P, Bignotto M, Del Papa N, et al. Antibody to carbonic anhydrase II is present in primary biliary cirrhosis (PBC) irrespective of antimitochondrial antibody status. Clin Exp Immunol 1998;114:448–54. PubMedCrossrefGoogle Scholar
Bizzaro N, Covini G, Rosina F, Muratori P, Tonutti E, Villalta D, et al. Overcoming a “portable” diagnosis in antimitochondrial antibody negative primary biliary cirrhosis: study of 100 sera and review of the literature. Clin Rev Allergy Immunol 2012;42:288–97. CrossrefGoogle Scholar
About the article
Published Online: 2017-07-21
Published in Print: 2017-11-27
Author contributions: All the authors have accepted responsibility for the entire content of this submitted manuscript and approved submission.
Research funding: None declared.
Employment or leadership: None declared.
Honorarium: None declared.
Competing interests: The funding organization(s) played no role in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; or in the decision to submit the report for publication.